Understanding Political Involvement Among Disadvantaged Adolescents

Transcription

1 Washington University in St. Louis Washington University Open Scholarship All Theses and Dissertations (ETDs) January 2009 Understanding Political Involvement Among Disadvantaged Adolescents Suzanne Pritzker Washington University in St. Louis Follow this and additional works at: Recommended Citation Pritzker, Suzanne, "Understanding Political Involvement Among Disadvantaged Adolescents" (2009). All Theses and Dissertations (ETDs) This Dissertation is brought to you for free and open access by Washington University Open Scholarship. It has been accepted for inclusion in All Theses and Dissertations (ETDs) by an authorized administrator of Washington University Open Scholarship. For more information, please contact

2 WASHINGTON UNIVERSITY George Warren Brown School of Social Work Dissertation Examination Committee: Amanda Moore McBride, Ph.D., Chair John Bricout, Ph.D. Nancy Morrow-Howell, Ph.D. Michael Sherraden, Ph.D. Edward Spitznagel, Ph.D. William Tate, Ph.D. UNDERSTANDING POLITICAL INVOLVEMENT AMONG DISADVANTAGED ADOLESCENTS by Suzanne Pritzker, MSW, MEd A dissertation presented to the Graduate School of Arts and Sciences of Washington University in partial fulfillment of the requirements for the degree of Doctor of Philosophy August 2009 St. Louis, Missouri

3 ABSTRACT OF THE DISSERTATION Understanding Political Involvement Among Disadvantaged Adolescents by Suzanne Pritzker Doctor of Philosophy in Social Work Washington University in St. Louis, 2009 Professor Amanda Moore McBride, Chairperson Following substantial concern in recent years about youth disengagement from the political arena, 2008 Presidential election data indicate that youth political participation is now on the rise. However, low-income and some ethnic minority youth are substantially less likely to participate and to hold positive attitudes about politics and government than their wealthier and/or White counterparts. This suggests a possible disconnect between ethnic minority or economically disadvantaged youth and the larger U.S. society, and may signal potential life-long disassociation from civic processes. Using four years of 12 th grade data from Monitoring the Future: A Continuing Study of American Youth, this dissertation sought to assess differential impacts of economic status and racial/ethnic identification on adolescent political development, to inform civic interventions that seek to counter this possible disconnect. Bivariate analyses tested previous findings of racial and socio-economic differences in political attitudes and behaviors. Confirmatory factor analyses assessed whether adolescents of different socio-economic status or races/ethnicities similarly interpret political attitude and behavioral measures. Finally, structural equation modeling analyses tested whether ii

4 socio-economic status or race/ethnicity moderate paths between political attitudes and behaviors. Results indicated that low-income, Black, and Hispanic adolescents held more negative political attitudes and lower levels of political behavior. While adolescents of different socio-economic status interpreted attitudinal and behavioral items and constructs similarly, important differences were identified across racial/ethnic groups. Specifically, Black adolescents interpreted political attitudes differently than other adolescents, suggesting caution in interpreting cross-group analyses of adolescent attitudes and perceptions towards government. Finally, path analyses indicated that White adolescents and adolescents across socio-economic status who are interested in government and hold political preferences and beliefs were more likely to engage in nontraditional political activities, while these attitudes did not predict electoral behavior. For these subgroups, positive views of how government acts predict electoral activity, while negative views of how government behaves predict non-traditional political activity. These paths were not all significant for Black, Hispanic, and Asian adolescents, however, suggesting that adolescent paths between political attitudes and behaviors may operate differently across races/ethnicities. Implications of these findings for youth development practice, future research, and civic education and civic development policy are presented. iii

5 ACKNOWLEDGMENTS This dissertation is the culmination of six years of study, guided by and supported by a great number of people to whom I owe my utmost gratitude. First and foremost, I would like to thank Professor Amanda Moore McBride, who has been an absolutely outstanding mentor throughout my doctoral studies. I have learned so much from Amanda about how to be a scholar, a teacher, and a mentor, and I am so grateful for her guidance, personal and professional support, enthusiasm, and the occasional much-needed pep talks! I also would like to thank each of the members of my dissertation committee: Professors John Bricout, Nancy Morrow-Howell, Michael Sherraden, Ed Spitznagel, and William Tate. The quality of my dissertation and my broader thinking about civic engagement across diverse communities benefitted substantially from the time they gave me and their thoughtful comments and critiques. Special thanks to fellow VCU alum John Bricout for his wonderful support and advice from the earliest stages of my doctoral education. Ed Spitznagel and Nancy Morrow-Howell were indispensable to me during my dissertation research, painstakingly helping me through my statistical analyses. None of this would have been possible without financial support from three sources at Washington University. The Center for Social Development, with funds from the Ford Foundation, generously supported both my dissertation research and other related research endeavors during my doctoral studies. The Mr. and Mrs. Spencer T. Olin Fellowship, under the leadership of Nancy Pope, has been invaluable to me, both in terms of financial support and in enabling me to make strong connections with fellow graduate iv

6 students across disciplines. The opportunity to be the inaugural recipient of the Birds of Passage Fellowship through the American Culture Studies certificate program helped support the final stages of my dissertation research. I am also appreciative of my colleagues and friends in the George Warren Brown School of Social Work community, from whom I have learned so much over the last six years. I hope that this is only the beginning of a lifetime of friendship and collaboration. I especially want to thank Paul Sacco for being so generous with his time and helping me to navigate Mplus, even with his own dissertation deadlines looming, and Jing Tan for sharing so many side-by-side hours of statistical analyses (and Ed) with me. To my friends and family both at Brown and beyond, I apologize for my frequent disappearances into dissertation-land and for often dragging Andy along with me, and am so thankful for your friendship. Finally, I want to thank my husband Andy for grinning and bearing through the extra house rehabbing load he took on, for staying up many late nights to give me moral support and critical feedback, and, most of all, for his never-ending patience and love. I am also very appreciative of my family for realizing that this was the right career choice for me, and for their love, support, and encouragement throughout. v

7 Table of Contents CHAPTER I: INTRODUCTION: AIMS AND SIGNIFICANCE... 1 Why Political Engagement?...3 Research Aims...7 CHAPTER II: RACE, SOCIO-ECONOMIC STATUS, AND ADOLESCENT POLITICAL PARTICIPATION... 9 Political Participation Among Racial and Ethnic Minority Youth...10 Political Participation Among Low-SES Youth...13 Disadvantage in Context...14 Institutional Context Marginalization and Group Identity Summary...18 CHAPTER III: THEORETICAL AND EMPIRICAL BACKGROUND: LINKING POLITICAL ATTITUDES WITH BEHAVIORS Adolescent Political Participation...20 Electoral Behaviors Political Voice Attitudinal Pathways to Political Participation...24 Social Attitudes Political Attitudes Summary...32 CHAPTER IV: THEORETICAL MODEL, RESEARCH QUESTIONS, AND HYPOTHESES Theoretical Model...36 Testing the Model: Research Questions and Hypotheses...38 Differences in Attitudes and Behaviors Across Subgroups Between-Group Differences in Interpreting Items Moderating an Attitude-Behavior Relationship CHAPTER V: RESEARCH METHODS Data and Sampling Procedures...45 Survey Description MTF Sampling Strategy Sampling Strategy for the Current Analysis Measures...48 Independent Variables Dependent Variables Moderating Variables Exogenous (Covariate) Variables Data Management and Analytic Methods...57 Sample Weights Bivariate Analyses Measurement Analyses Structural Analyses Missing Data Analysis and Multiple Imputation CHAPTER VI: RESULTS vi

8 Sample Description...76 Bivariate Findings...78 Political Attitudes Political Behaviors Measurement Findings...84 Measurement Invariance Across Income Models Measurement Invariance Across Race/Ethnicity Models Structural Findings Multiple Group Analyses Across Income Models Multiple Group Analyses Across Race/Ethnicity Models CHAPTER VII: DISCUSSION Summary and Analysis of Findings Research Question Research Question Research Question Limitations Theoretical Model Measurement Methodology Implications Implications for Practice Implications for Research Implications for Policy Conclusion REFERENCES APPENDIX vii

9 List of Tables Table 5-1: MTF survey 12 th grade sample Table 5-3: Missing data analysis (percent missing), by race and income Table 5-4: Missing data analysis (percent missing), by gender, age, and geographic residence Table 5-5: Additional MTF variables used to predict imputed data Table 6-1: Sample demographics Table 6-2: Demographic characteristics by race/ethnicity group Table 6-3: Demographic characteristics by income group Table 6-4: Differences in categorical political attitudes and behaviors, by income status 80 Table 6-5: Differences in continuous political attitudes, by income status Table 6-6: Differences in categorical political attitudes and behaviors, by race Table 6-7: Differences in continuous political attitudes, by race Table 6-8: Goodness-of-fit of the baseline measurement model across subgroups Table 6-9: Goodness-of-fit statistics for models testing invariance across income Table 6-10: Changes in goodness-of-fit statistics across levels of invariance, by income 90 Table 6-11: Model fit statistics for models testing invariance across race Table 6-12: Changes in model fit statistics across levels of invariance, by race Table 6-13: Factor loadings with differences between race subgroups Table 6-14: Full structural model with covariates Table 6-15: Model fit of the baseline structural model across subgroups Table 6-16: Path coefficients, by income Table 6-17: Path coefficients, by racial/ethnic group viii

10 List of Figures Figure 4-1: Theoretical model: The effect of disadvantage on the relationship between adolescent political attitudes and behaviors Figure 4-2: Theoretical model with observed variables Figure 5-1: Measurement model Figure 5-2: Structural model Figure 6-2: Low-income measurement model Figure 6-3: Non-low-income measurement model Figure 6-4: Measurement model for White subsample Figure 6-5: Measurement model for Black subsample Figure 6-6: Measurement model for Hispanic subsample Figure 6-7: Measurement model for Asian subsample Figure 6-8: Structural model with no covariates Figure 6-9: Structural model with covariates Figure 6-10: Low-income structural model Figure 6-11: Non-low-income structural model Figure 6-12: Structural model for Black subsample Figure 6-13: Structural model for White subsample Figure 6-14: Structural model for Hispanic subsample ix

11 CHAPTER I: INTRODUCTION: AIMS AND SIGNIFICANCE The 2008 presidential campaign defied previous scholarly and conventional wisdom about youth participation in politics. In the November 2008 general election, approximately 23 million youth ages voted. This reflects an increase of about 4-5% over the 2004 general election youth turnout and at least 11% over the 2000 turnout. This increased youth vote accounted for over 60% of the overall national voter turnout increase (Center for Information and Research on Civic Learning and Engagement [CIRCLE], 2008c). Similar trends were apparent during the primary season. In all but one of the 17 primary or caucus states 1 for which a comparison year is available, youth ages voted at higher rates than in 2000, in some cases tripling or quadrupling turnout. Rates of youth participation ranged from a low of 5% in Nevada s caucuses to 43% in New Hampshire s primary (CIRCLE, 2008a; Kirby, Marcelo, Gillerman, & Linkins, 2008). Youth interest in the 2008 campaigns was evident not just in increased voting rates, but also in the substantial attention to candidates and the campaign on internet social networking sites young people frequent such as My Space and Facebook. This growing political attention counters the widespread notion that today s youth are disengaged from the political arena (e.g., Galston, 2004; Keeter, Zukin, Andolina, & Jenkins, 2002a). Yet, despite these positive youth participation trends, questions remain. Who are these youth participants, or perhaps more importantly, who are the non-participants? Analyses of the overall racial and ethnic or socio-economic demographics of this increased youth participation have not yet been conducted. However, if previous trends 1 The one exception is New York, where youth participation remained steady. Interestingly, contrary to the pattern in other states, overall primary voting in New York was lower in 2008 than in 2000 (CIRCLE, 2008). 1

12 are maintained, then we can expect to see that White youth vote more than non-whites, and that Black youth vote more than any other minority group (Lopez & Kirby, 2005; Marcelo, Lopez, & Kirby, 2007) 2. Lower-income youth similarly may vote at lower rates than wealthier youth (Lopez, Kirby, Sagoff, & Kolaczkowski, 2005). Such disparities are not limited to voting. Racial and ethnic differences have been found across an array of political attitudes and behaviors 3 (Lopez & Kirby, 2005; Lopez, et al., 2006a), and socio-economic status and educational levels have been found to significantly predict civic knowledge and participation (Corporation for National and Community Service, 2006; Torney-Purta, et al., 2001). These data not only suggest a possible disconnect between youth who are ethnic minorities or economically disadvantaged and the larger U.S. society, but also signal potential life-long disassociation from civic processes. Adolescence may be a particularly crucial time for intervention to counter this disconnect. Theoretical links between adolescent engagement with politics and adult political behavior have long been an element of political science thought (Berti, 2005; Dudley & Gitelson, 2003; Searing, Schwartz, & Lind, 1973; Searing, Wright, & Rabinowitz, 1976), with substantial support for claims that early patterns of political behavior are repeated into adulthood (Hooghe, 2004). Developmental research also emphasizes the role of adolescent civic experiences in shaping adult political engagement (Flanagan, 2003; Youniss, McLellan, & Yates, 1997). The development of civic identity is understood to be a key task of early and 2 In the 2004 Presidential election, for example, among year olds, voting rates were as follows: 49.8% of white youth, 47.3% of black youth, 36.6% of Native-American youth, 35.5% of Asian-American youth, and 33.0% of Latino youth (Lopez & Kirby, 2005). 3 Generally, citizens are understood to behave politically when they participate in activities through which they can make their wishes known to politicians and government officials (Verba, Schlozman, & Brady, 1995). This goes beyond voting to include a wide range of behaviors that directly or indirectly affect government action. Additional political behaviors are identified and discussed in depth in Chapter 3. 2

13 middle adolescence (Erikson, 1968; Sherrod, 2006; Youniss & Yates, 1997). Through learning, observation, and participation in various civic experiences 4, adolescents are expected to absorb messages about the civic activity required of them as citizens, developing attitudes and values that can shape their civic behaviors both in adolescence and into adulthood (Flanagan, 2003; Syvertsen & Flanagan, 2005; Youniss, et al., 1997). One aspect of civic identity adolescents may develop consists of attitudes towards politics and government. In this study, links between these political attitudes and adolescents political behavior are examined. It is posited that the experience of belonging to a group that historically has been disadvantaged in terms of political processes impacts the process by which adolescents become political actors (e.g., Sanchez-Jankowski, 2002). By understanding more about relationships between attitudes and behaviors and how these may be affected by disadvantage or marginalization, we can better design interventions to target relevant outcomes in adolescent participants. In this introductory chapter, the importance of this research to the broader policy arena is presented, and the aims of the dissertation are described. Why Political Engagement? Participation in political activities is an important way that citizens can affect the circumstances in which they live and the resources to which they have access. It entails making viewpoints heard in such a way that politicians and others involved in policymaking can become aware of these perspectives and transfer them into policy. Citizens who participate politically can make their voices heard on issues close to home such as the location of neighborhood bus stops or placement of landfills, and on broader issues 4 The development of civic identity also may be influenced by other sources such as family and peers. 3

14 with national and international impact such as whether funding is needed for social service programs or military activities. Widespread citizen engagement with a democratic polity may result in policies that are responsive to the interests and needs of a diverse national population. When participation is unequal across the populus, however, policies that are unresponsive, perhaps socially unjust, may result. For example, although some scholars have found little difference between non-voters and voters on certain policy issues (Bennett & Resnick, 1990; Teixera, 1992, as cited in Lijphart, 1997); others have found that class inequalities in political participation can impact policy outcomes. For example, studies analyzing relationships between voter turnout and tax and welfare policy outcomes have found that states with a disproportionately wealthy electorate grant lower welfare benefits to their citizens (Hill & Leighley, 1992; Hill, Leighley, & Hinton-Anderson, 1995), thus demonstrating a possible negative relationship between non-participation by low-income citizens and access to needed services. Although non-voters and voters may share similar opinions on a given set of policy issues, the issues that they would prefer policy-makers to address may differ. Verba, et al. (1997) found that economically disadvantaged citizens were more likely to discuss issues related to basic human needs, while those with more resources were more interested in discussing economic or social issues 5. Disadvantaged adults have been found to send to public officials on average fewer than one-fourth of the number of political messages sent by more advantaged adults (Verba, et al., 1997). Accordingly, politicians may be more likely to focus attention on issues of concern to advantaged 5 Verba, et al. (1997) make these conclusions based on a comparison between citizens with no education beyond high school and family incomes below $20,000 in 1990, and citizens with at least some college education and family incomes above $50,000. 4

15 citizens, leaving the most disadvantaged citizens with institutions insufficiently responsive to their interests and needs (Verba, et al., 1997). Race and ethnicity, too, have been linked to differing policy preferences. For example, African-American and Latino adults may be more likely than Whites to be interested in promoting such issues as those involving civil rights, crime, and the needs of children and youth (Schlozman, Page, Verba, & Fiorina, 2004). On an intuitive level, one might expect that people with a direct, personal stake in policy outcomes would be more likely to participate in political processes (Rosenstone & Hansen, 1993); however, many of those who are direct recipients of government social welfare policies are not political participants. Accordingly, they are less likely to have their political concerns communicated to public officials than more advantaged citizens (Verba, et al., 1997). Non-participation in the political sphere may have individual-level impacts as well. Citizens who do not participate in the civic lives of their community or nation may perceive that they have fewer rights than others. They may feel disempowered and disengaged from the larger society, perhaps less concerned with preserving the well-being of their communities than other citizens. Adolescent political engagement may be not only an avenue for countering this adult non-participation 6 ; it also may serve purposes intrinsic to adolescence. The scholarship of Barry Checkoway and colleagues is particularly influential in this regard. 6 It is important to clearly note that non-participation cannot be overcome just through increased attention to adolescent civic development. Rather, institutional structures may limit participation among disadvantaged groups. For example, although such policies are increasingly changing across the nation, one-day on-site voting or caucus structures may exclude low-income individuals working multiple jobs or with inflexible work schedules. Recent legislation calling for the use of voter identification, particularly photo identification, in some states has been found to exclude minority citizens, immigrants, and lowincome citizens from voting (Barreto, Nuño, & Sanchez, 2007). The declining levels of civic education in schools (CIRCLE, 2003; U.S. Department of Education, 2001) may also contribute to non-participation, as citizens may not feel they have sufficient knowledge to participate in political activities. 5

16 Checkoway, et al. (2003) argue that youth are competent citizens who can be assets to their communities and critical players in civic life. This perspective suggests that youth are legitimate political actors in their own right, who have the capacity to affect the environment in which they live (Alparone & Rissotto, 2001) and to engage in public decision-making (O'Donoghue, Kirshner, & McLaughlin, 2002). It is expected that through participation in political activities, youth can both gain civic skills and improve the well-being of their communities (Finn & Checkoway, 1998). An emphasis on adolescent political participation not just as a predictor of adult participation but also as a desired outcome in and of itself is relatively new in civic engagement scholarship. However, it is consistent with international scholarship and policy work, including the 1989 United Nations Convention on the Rights of the Child 7, that has sought recognition of children and adolescents under 18 as citizens with agency (Sapiro, 2004). Grassroots attention in some parts of the U.S. also has called attention to political participation during adolescence. Some activists have lobbied for the reduction of the voting age, and legislation to do so has been considered in a number of U.S. localities and states (National Youth Rights Association, 2007). Before proceeding with this analysis, however, it should be noted that increased political engagement is not universally supported. For example, a subset of political theorists have long argued that expansive political participation could bring about democratic instability (Pateman, 1970). Such arguments suggest that average, i.e., nonelite, citizens do not have the interest, realistic perspectives, or expertise necessary to impact policy decisions (Pateman, 1970). Is it possible that less-educated, less politically-savvy citizens could be particularly susceptible to bad influences of 7 Interestingly, only two countries -- the United States and Somalia -- have not ratified the CRC. 6

17 particular politicians or political movements? Mid-20 th century political theorists like Robert Dahl certainly thought so, arguing that increased participation among low-income Americans could threaten democracy due to their supposed authoritarian personalities (Pateman, 1970). Fiorina (1999) offers a different critique of increased political participation, that higher levels of political participation can lead to more negative policy outcomes. He argues that in previous eras, when U.S. citizens were less involved in political processes, citizens were happier with government. As policy and politics has become more accessible to the citizenry as a whole, however, Fiorina (1999) believes that this has led to political polarization, and to the disproportionate power of extreme voices relative to the moderate views of the majority of Americans. Research Aims The ultimate aim of this dissertation research is to inform how social workers, educators, and other professionals who work with adolescents can increase political behaviors on the part of low-income and racial and ethnic minority adolescents. While aspects of disadvantage may reduce the civic resources accessible for engagement (Verba, et al., 1995), civic interventions may counter these negative impacts of disadvantage on youth civic engagement (Checkoway, et al., 2003; Cohen, 2006; Ginwright, 2006; Kirshner, Strobel, & Fernandez, 2003; Sherrod, Flanagan, & Youniss, 2002; Watts & Guessous, 2006). Of interest then is strengthening our knowledge base in order to design civic interventions to most effectively meet goals of increasing political participation. Interventions such as service-learning, extracurricular activities, and classroom civic education have shown some potential for increasing civic identity and the 7

18 corresponding attitudes that may lead to future political behavior (Flanagan, 2003; Pritzker & McBride, 2006b; Yates & Youniss, 1998). To inform effective interventions for diverse groups of youth, it is necessary first to identify variables that can be targeted and manipulated in order to elicit increased political participation. Given the emphasis on civic identity in youth development research (e.g., Sherrod, et al., 2002; Youniss & Yates, 1997), this dissertation is interested in mutable outcomes related to civic identity; that is, how youth see themselves in relation to the broader society and polity (Flanagan, 2003; Youniss, et al., 1997). An analysis of civic interventions finds that interventions regularly seek to impact aspects of civic identity including social attitudes such as social trust and social responsibility; and political attitudes such as interest in government, political efficacy, and perceptions of government responsiveness (Pritzker & McBride, 2006b). Increased understanding of relationships between these attitudes and different forms of adolescent political participation will help inform modification of interventions to better target those aspects of identity most likely to further political participation. This dissertation also assesses whether these relationships differ across race, ethnicity, and class, to inform civic programming specifically with minority and low-ses adolescents. Specific aims are as follows: 1. To test relationships between political attitudes and political behaviors among adolescents. 2. To assess differential impacts, based on economic status and racial/ethnic identification, of political attitudes on political behaviors. 8

19 CHAPTER II: RACE, SOCIO-ECONOMIC STATUS, AND ADOLESCENT POLITICAL PARTICIPATION As noted in Chapter 1, eschewal of political involvement tends to be prevalent among low-income and minority adolescents. The participation of diverse youth in political processes relevant in its own right is increasingly worthy of attention as income inequalities continue to increase and demographic shifts in the United States result in growing representation of non-whites and immigrants among the youth population. Among 18 to 25 year olds, for example, the percentage of Whites has dropped, while the percentages of African-American, Hispanic, and immigrant youth are growing 8 (Corporation for National and Community Service, 2006; Lopez & Marcelo, 2006; Marcelo & Lopez, 2006). Given the apparent marginalization of some low-income and minority adolescents from political processes, effective interventions to increase political participation should take into account ways in which disadvantage may affect political engagement. Although this analysis places particular emphasis on ethnic and economic disadvantage, it should be noted that other adolescent circumstances also may negatively impact political engagement. For example, the very nature of being an adolescent may be politically marginalizing, as those under the age of 18 are unable to participate in certain forms of electoral activity; females also may feel outside of political processes (Jenkins, 2005; Taft, 2006), as was commonly discussed during the 2008 Democratic primary campaign. Additionally, youth with disabilities (Skelton & Valentine, 2003), homosexual youth 8 The percentage of youth ages that is non-white has increased from 22% to 28% over the last 30 years (Marcelo, et al., 2007). 9

20 (Russell, 2002), and immigrant youth (Stepick & Stepick, 2002) may all find themselves at times outside of mainstream political processes in the United States. The context for this dissertation study is established in this chapter through an analysis of current knowledge about political engagement among low-income, minority adolescents. It is important to acknowledge that race and socio-economic status in and of themselves likely are insufficient to explain these adolescents political disengagement. Rather, low-income and minority adolescents may be disproportionately likely to live in communities where few civic resources are available and to experience marginalization or discrimination that may leave them feeling excluded from political processes. Key contextual factors that may contribute to their generally low rates of political participation and may limit their access to civic interventions are discussed. Political Participation Among Racial and Ethnic Minority Youth When investigating racial and ethnic differences, political participation scholarship has typically focused exclusively on White and Black citizens (Leighley & Vedlitz, 1999; Marschall, 2001). As a result, much less is known about political behaviors among non-black minorities. For example, Torney-Purta, et al. (2007) note that there is particular lack of understanding about the political development of Hispanic youth. Likewise, Asian-American youth are rarely included in surveys of youth civic engagement (Lopez, et al., 2006a). When political engagement has been compared across youth of different races and ethnicities, some differences consistently have been identified. Although Black youth ages report greater involvement than other ethnic groups in an array of political activities such as voting, donating money to campaigns, canvassing for a candidate, and 10

21 contacting media (Lopez, et al., 2006a), White youth were most likely to vote in the 2004 Presidential election 9 (Lopez & Kirby, 2005). Political disengagement is particularly prevalent among Hispanic year-olds, who report the lowest levels of participation in political activities such as signing petitions, participation in political groups, and boycotting (Lopez, et al., 2006a; Torney-Purta, Barber, & Wilkenfeld, 2006; Torney- Purta, et al., 2007). On the other hand, Hispanic and immigrant youth and youth who are children of immigrants report greater involvement in political protests than other youth (Lopez, et al., 2006a). White youth generally appear to hold more positive attitudes toward politics and government than Black youth 10. When other ethnic groups are included in analyses, they also exhibit less positive attitudes regarding politics than Whites, and often than Black youth as well. This pattern is present even as young as age 14 (Fridkin, et al., 2006). While 57% of White youth consider voting an important activity, less than half of African-American and Hispanic youth do (Lopez & Kirby, 2005). African-American and Hispanic youth also are less likely than Whites to feel they can make a difference in their communities (Lopez & Kirby, 2005). Data indicate particularly low trust in government among Blacks, Hispanics, and Native-Americans, with minority youth more likely than 9 All racial and ethnic groups and non-college attending youth increased their rate of voting between the 2000 and 2004 Presidential elections (Lopez & Kirby, 2005). This, plus the overall increases in youth voting in the 2006 midterm election may signal a growing interest among youth across populations in political participation. Youth turnout data by racial groups has not yet been published for the 2008 election; however, preliminary analyses suggest that Hispanic and Black youth voted at a greater rate than their distribution in the overall electorate (CIRCLE, 2008) 10 Likewise, Black and Latino adults historically have exhibited more negative views about politics and government than White adults (e.g., Guterbock & London, 1983; Shingles, 1981; Tate, 1991; Verba, et al., 1995). 11

22 Whites to feel government excludes them 11 (Flanagan & Faison, 2001; Fridkin, et al., 2006; Lopez, 2003), although Baldi, Perie, Skidmore, Greenberg, and Hahn (2001) found insignificant differences in trust between ethnic minority youth. Where Asian-American youth have been included in studies, though, they appear more likely than other minorities to view the political system as responsive (Lopez, et al., 2006a). For Hispanics, low levels of engagement may be attributable in part to immigrant status, due either to being an immigrant or to the recency of family immigration (Stepick & Stepick, 2002; Torney-Purta, et al., 2006). Data on civic knowledge also reflect these racial and ethnic patterns, with Whites reporting the highest average civic knowledge scores at 4 th, 8 th, and 12 th grades on the National Assessment of Educational Progress [NAEP] study of civic competency. A substantial knowledge gap exists between White students and Black and Hispanic students. Though narrower for 4 th graders than when the study was previously administered in 1998, the gap between White and Hispanic students at all three grade levels is at least 24 points on a 300-point scale. The gap between White and Black students, statistically unchanged since 1998, is at least 25 points at all three grade levels (Lutkus & Weiss, 2007). Among 9 th graders, Baldi, et al. (2001) found White youth scored higher than Black and Hispanic students in terms of civic knowledge and skills; Asians consistently exhibited higher knowledge and skills than Black students, and in some cases, than Hispanic students as well. Other studies suggest that civic knowledge may be particularly low among Hispanic adolescents compared to non-hispanics (Torney-Purta, et al., 2007). 11 It remains to be seen whether the election of the first minority U.S. president will impact any of these attitudinal dynamics among minority youth. 12

23 Political Participation Among Low-SES Youth Less is known about differential patterns of political engagement across socioeconomic status. With educational attainment as a proxy for SES, non college-attending year olds are less likely than those who have attended college to vote, to identify voting as important, or to discuss politics with their parents (Lopez, et al., 2005). A socio-economic gap also seems evident in terms of the civic knowledge that may precede political participation. Across 4 th, 8 th, and 12 th grades, students eligible for free or reduced school lunches, another proxy for poverty, had lower civic knowledge scores than wealthier students on the NAEP study of civic knowledge (Lutkus & Weiss, 2007) 12. A similar civic knowledge gap and lower willingness to vote was found among U.S. 14- year olds living in high-poverty areas with few books in their home and low educational aspirations, as compared to other American students (Torney-Purta, 2001). Among adults, data indicate class gaps in terms of political behaviors and attitudes (Verba, et al., 1995). For example, Verba, et al. (1995) find a strong positive relationship between income and internal political efficacy. There is some debate as to whether the availability of resources made possible by one s socio-economic class may provide a stronger explanation for differences in political behaviors than the racial differences described in the previous section (Verba, Schlozman, Brady, & Nie, 1993a). In a nutshell, Verba s resource theory suggests that the resources time, money, and civic skills believed to be necessary for political participation are more likely to be present among individuals of higher socio-economic 12 Among 8 th graders, for example, 48% of students qualifying for free lunches met the test s basic level, as compared to 82% of students too wealthy to be eligible for free or reduced lunches. Among 12 th graders, 35% of students whose parents did not graduate high school met the basic level, in contrast with 77% of students with at least one college graduate parent (Lutkus & Weiss, 2007). 13

24 status (Brady, Verba, & Schlozman, 1995; Verba, et al., 1995; Verba, et al., 1993a). The effects of low socio-economic status are expected to accumulate, resulting in low levels of political participation among adults with low levels of education and income (Marschall, 2000; Verba, et al., 1997). In fact, when SES is taken into account, some studies have found that the racial differences described above do not hold; instead, Black adults may participate in political activities at a higher rate than Whites (Marschall, 2001) and Hispanics may vote at approximately the same rate as the general population (Michelson, 2000). Adolescent participation may well be impacted by family resources (Fridkin, et al., 2006; Verba, Burns, & Schlozman, 2003). Educated parents may be better equipped to transmit political skills to their children, and parental income may increase the likelihood of political information being available to adolescents in their homes. However, as noted previously, little is known about the degree to which civic attitudes and behaviors differ for adolescents across income levels. Disadvantage in Context Although race and socio-economic status are consistently linked to political participation, they do not exist in isolation, and may be insufficient in and of themselves to explain differences in political participation (Piven & Cloward, 1988). Rather, there are broader group, community, and societal factors that may serve to increase lowincome and minority adolescents marginalization from political processes. This may be especially relevant in the case of political behaviors which, other than the vote, often do not take place in isolation (Frasure, 2003). In the following sections, aspects of institutional context that may limit the civic resources available to disadvantaged 14

25 adolescents and factors that may shape how low-income and minority adolescents experience political processes are discussed. Institutional Context The community environment can serve as a political stimulus or as a barrier to political participation. For example, Gimpel, et al. (2003) suggest that minority youth who live in heterogenous neighborhoods in proximity to other racial or ethnic groups are more likely to discuss politics and be politically efficacious, although Putnam (2007) finds that living in areas of greater diversity may be associated with lower levels of trust and civic participation. Of particular relevance for disadvantaged adolescents is how neighborhood poverty may impact political engagement for low-income or minority adolescents. Cohen (2006) and Ginwright (2006) argue that adolescents exposed to poverty and concomitant social problems may engage politically in ways different from other youth. Interactions with government entities such as public education systems and the police may yield strong opinions on the part of adolescents about their communities, government, and the political system (Cohen, 2006; Ginwright, 2006). High ratios of adolescents to adults (Hart, Atkins, Markey, & Youniss, 2004) and a lack of social interactions (Bolland & McCallum, 2002; Lay, 2006) in poor urban neighborhoods may contribute to low civic knowledge and participation. Interestingly, low-income rural adolescents exhibit higher levels of political knowledge than low-income urban adolescents (Lay, 2006). High-poverty urban communities may lack institutions that can connect citizens with the larger community or polity (Kirshner, et al., 2003; Marschall, 2000). 15

26 Community institutions offer opportunities for mobilizing citizens into civic activity or for transmitting skills that are key to political activity (Verba, et al., 2003), but may be more available to advantaged adolescents (Hart, Atkins, & Ford, 1998). With few civic institutions, youth living in poverty may have insufficient support for civic and political development, limited opportunities to develop civic competence, and few adult role models who are actively engaged with the community or the polity (CIRCLE, 2003; Flanagan & Faison, 2001; Hart & Atkins, 2002; Kirshner, et al., 2003). Urban schools in high-poverty areas may be particularly likely to lack the financial resources and time to devote to the development of civic and democratic skills that may exist in more advantaged communities (Hart & Atkins, 2002; O'Donoghue & Kirshner, 2003), contributing to a civic achievement gap (Levinson, 2007). Schools serving poor and working-class minority youth may even educate students away from these obligations of citizenship and toward civic alienation (Fine, Burns, Payne, & Torre, 2004, p. 2212). However, while some research finds that urban schools offer fewer civic development opportunities (O'Donoghue & Kirshner, 2003), a recent national survey of public school principals indicates that high-poverty urban schools may be particularly likely to embrace service-learning, a potential tool for increasing civic engagement (Pritzker & McBride, 2006a). Marginalization and Group Identity Racial or ethnic group identity and experience with societal discrimination and marginalization also may impact minority adolescent political participation (e.g., Fridkin, et al., 2006). Among adolescents who are racial and ethnic minorities and have limited access to financial resources, some young people already feel that they are on the 16

27 margins of democratic life (Fridkin, et al., 2006, p. 606) or that government acts against them in some way (Bedolla, 2000; Taft, 2006). The data presented earlier in this chapter provides some evidence for this argument, finding lower participation and more negative attitudes toward government among adolescents who are low-income or members of certain minority groups. Sanchez-Jankowski (2002) suggests that group histories of racial exclusion in the United States may play a central role in shaping how minority adolescents respond to civic institutions and engage with their communities. Minority youth may receive conflicting messages about civic engagement: while formal institutions, such as schools, media, and the government, prioritize traditional forms of civic and political participation, local informal institutions may encourage group-focused forms of engagement. Because of historical exclusion from formal institutions, Sanchez- Jankowski (2002) suggests that some minority youth may be more likely to respond to civic messages from local, informal institutions and thus engage primarily in efforts to help their own communities. This theory seems to play out in terms of some Black youth who lack faith in the broader political system and instead actively advocate for quality of life issues in their own communities (Ginwright, 2006) and working-class Hispanic youth who may prioritize solving local community problems over engaging in traditional political processes (Bedolla, 2000). Sanchez-Jankowski (2002) notes, however, that some Asian youth may engage with civic institutions differently than Blacks and Hispanics; for example, they may hold more positive attitudes about the national government. He attributes this to an experience of racial inclusion in the U.S., in 17

28 which Asians may have been discriminated against initially, but now are fully integrated into society. Marginalization may shape political participation in other ways as well. With their social domination theory, Sidanius, Pratto, van Laar and Levin (2004) suggest that youth experiences of civic socialization are affected by membership in a subordinate group. A national study found that as Black and White adolescents took more civics classes in school, they increasingly defined citizenship differently from each other, mirroring societal structures (Sidanius, et al., 2004). Marginalization may also contribute to feelings of stigma that can impact how individuals perceive politics, thus determining whether they become politically alienated or politically active (Schur, Shields, & Schriner, 2003). Summary Adolescents who are low-income or belong to a racial or ethnic minority group are less likely than other youth to engage in political activities or to possess the kinds of knowledge and attitudes that encourage future participation. Those living in high-poverty communities face a particular deficit in exposure to opportunities for civic socialization that could promote their engagement. Given these predilections away from political engagement, it is important to examine how political participation can be engendered among low-income and minority adolescents. In the next chapter, ways that race, ethnicity, and socio-economic status can shape how adolescents become politically engaged are examined. 18

29 CHAPTER III: THEORETICAL AND EMPIRICAL BACKGROUND: LINKING POLITICAL ATTITUDES WITH BEHAVIORS The aspects of disadvantage discussed in Chapter 2 appear influential in terms of who participates politically and who does not. Among those low-income and minority adolescents who data suggest may be less likely to participate, how can political participation be increased? Through civic activities, adolescents can develop political knowledge and skills and a civic identity that can lead to political participation (Flanagan, 2003; Sherrod, et al., 2002; Youniss, et al., 1997). Political knowledge specifically refers to literacy regarding politics, current events, and processes of bringing about political change, as well as the capacity to make knowledgeable political choices. Knowledge is closely tied to political skills, which are capabilities to engage in democratic processes, such as the ability to engage constructively in processes of debate and negotiation. Empirical findings of association between measures of adolescent political knowledge and skills and political participation are fairly common (Galston, 2001; Kirlin, 2003; Verba, et al., 1995). But do particular attitudinal elements of identity also have strong links with political participation? In order to strengthen civic interventions, we must first identify whether particular aspects of civic identity do in fact precede political participation, and the extent to which this is the case for low-income and ethnic minority adolescents. In this chapter, the theoretical and empirical work that informs this study is examined, focusing primarily on potential attitudinal predictors of adolescent political behaviors. The chapter begins with an examination of the set of behaviors that constitute adolescent political participation. Particular attention is paid to scholarly distinctions 19

30 between participation that is electoral and that defined as political voice activity. Then, attitudinal factors that are commonly theorized to precede political participation are explored. Links between civic attitudes and participation are discussed in regard to adolescents in general and specifically in regard to low-income and minority adolescents. Adolescent Political Participation Political participation is one component of civic engagement, a broad concept that encompasses a multitude of ways that citizens can connect with the larger society. Civic engagement may refer to both social and political forms of activity (McBride, 2003). Broadly, social forms of engagement are connected to a general public good outside the policy arena, while political forms are connected to politics and policy (McBride, 2003). Although disagreements over the boundaries of what is considered civic engagement are widespread in the scholarship (Gibson, 2001; Obradovic & Masten, 2007; Sherrod, et al., 2002; Westheimer & Kahne, 2004; Youniss, et al., 2002), scholars and public officials generally share a concern over the degree to which youth do or do not connect with the larger society. This dissertation research is limited to those behaviors that take place directly in the political sphere, where concern over low levels of engagement has been particularly prevalent. This is because of the close relationship between political activity and possible policy change. As noted in Chapter 1, the policy consequences of non-engagement in political activities may be substantial. This is a particular concern in regard to today s adolescents, for whom there has been some evidence that socially-oriented engagement is a substitute for, rather than a complement to, political activity (Galston, 2001; Theiss- Morse & Hibbing, 2005; Walker, 2002). 20